From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <49E982C4.8020407@cosmosbay.com>
Date: Sat, 18 Apr 2009 09:35:32 +0200
From: Eric Dumazet <dada1@cosmosbay.com>
MIME-Version: 1.0
References: <20090313183303.GF3436@psychotron.englab.brq.redhat.com>
	<20090415081720.GA21342@psychotron.englab.brq.redhat.com>
	<20090415081819.GB21342@psychotron.englab.brq.redhat.com>
	<20090415180215.GA22540@psychotron.englab.brq.redhat.com>
	<20090417115723.GE9556@psychotron.englab.brq.redhat.com>
	<20090417083315.2f089755@nehalam>
	<20090418070151.GA3370@psychotron.englab.brq.redhat.com>
In-Reply-To: <20090418070151.GA3370@psychotron.englab.brq.redhat.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Subject: Re: [Bridge] [PATCH 1/3] net: introduce a list of device
 addresses	dev_addr_list (v3)
List-Id: Linux Ethernet Bridging <bridge.lists.linux-foundation.org>
List-Unsubscribe: <https://lists.linux-foundation.org/mailman/listinfo/bridge>, 
	<mailto:bridge-request@lists.linux-foundation.org?subject=unsubscribe>
List-Archive: <http://lists.linux-foundation.org/pipermail/bridge>
List-Post: <mailto:bridge@lists.linux-foundation.org>
List-Help: <mailto:bridge-request@lists.linux-foundation.org?subject=help>
List-Subscribe: <https://lists.linux-foundation.org/mailman/listinfo/bridge>, 
	<mailto:bridge-request@lists.linux-foundation.org?subject=subscribe>
To: Jiri Pirko <jpirko@redhat.com>
Cc: ivecera@redhat.com, fubar@us.ibm.com, netdev@vger.kernel.org, bridge@lists.linux-foundation.org, linux-kernel@vger.kernel.org, mschmidt@redhat.com, bonding-devel@lists.sourceforge.net, Stephen Hemminger <shemminger@vyatta.com>, jgarzik@pobox.com, davem@davemloft.net

Jiri Pirko a =E9crit :
> Fri, Apr 17, 2009 at 05:33:15PM CEST, shemminger@vyatta.com wrote:
>=20
> <snip>
>=20
>>> +struct netdev_hw_addr {
>>> +	struct list_head	list;
>>> +	unsigned char		addr[MAX_ADDR_LEN];
>>> +	int			refcount;
>>> +	struct rcu_head		rcu_head;
>>> +};
>> Minor nit, the ordering of elements cause holes that might not be
>> needed.
>=20
> Agree that ordering might be done better. Will do.
>> Space saving? is rcu_head needed or would using synchronize_net
>> make code cleaner and save space.=20
>>
>=20
> Well I originaly had this done by synchronize_rcu(). Eric argued that i=
t might
> cause especially __hw_addr_del_multiple_ii() to run long and suggested =
to use
> call_rcu() instead. I plan to switch this to kfree_rcu() (or whatever i=
t's
> called) once it hits the tree.
>=20

Yes, and dont forget we wont save space, as we allocate a full
cache line to hold a 'struct netdev_hw_addr', since we dont want this
critical and read_mostly object polluted by a hot spot elsewhere in kerne=
l...

Considering this, letting 'rcu_head' at the end of structure, even if we
have an eventual hole on 64 bit arches is not really a problem, and IMHO
the best thing to do, as rcu_head is only used at dismantle time.

And yes, maybe kfree_rcu() will makes its way in kernel, eventually :)

Thank you